Re: HTML Streaming (Chris Maden) wrote:

>Font metrics are determined by the *user's preferences*.  The
>served-out document can not describe that.  There is no way to know
>that your paragraph is going to take up 12 pixels on my really-wide
>Lynx display, or that the same paragraph will take up 7 vertical
>inches on my nearsighted granny's WebTV display.  Or that it will take
>30 seconds on my buddy Phil's screen reader.  The server has no
>meaningful way of describing the physical extent of a textual

The descriptions, added to the EXTENT tag, would work in concert with the
browser preferences. The HTML editor describes the HTML file. It is then up
to the browser to use the information as a hint to pre rendering the page.

>Please show, in a concrete example, how this document or its HTTP
>headers would change, and how a browser's rendering behavior could be
>expected to change in response.

<event p=200 table=75,25>
Imagine 200 character and spaces here
Imagine a table that is 75 by 25

The browser would first display a pre rendered page for a paragraph with 200 
character and spaces and then a table that is 75 by 25. The browser would 
then stream the character of the paragraph and cells of the table. Currently,
the entire paragraph would have to be downloaded first. Then the entire
table before it could be displayed.

The problem could be solved with the addition of extent tag to the paragraph
tag. This is what I am trying to avoid. Many other tags and new tags do not
have pre rendering attributes. Unless the entire page is described, the page 
cannot be pre rendered. For example, the table tag already has a pre 
rendering attribute, colgroup. But it cannot be displayed until the 
paragraph is completely downloaded because the paragraph does not have a pre 
rendering attribute.

I don't think every tag that does not have a pre rendering attribute should
be given one. I don't think that every new tag should be given a pre 
rendering attribute. I think their should be an extent tag with general 
description for size attributes, information etc.

Albert Fine

Received on Monday, 1 September 1997 09:45:29 UTC